Current genomic methods are limited to reporting an average signal from a complex population of cells because they require a large amount of input material. In complex tissues such as tumors, where genetic heterogeneity is common, important information may be lost. To address this problem, we propose to develop a whole-genome sequencing method that can obtain high-coverage (>80%) sequencing data from the genome of a single human cell. From this data we will identify the full spectrum of somatic mutations, including point mutations, indels, structural variants and copy number aberration that are present in the genome of a single cell. By comparing the genomes of multiple cells we can delineate clonal diversity and infer how tumor genomes evolve complex somatic mutations. To do this, we propose to develop an innovative method called Cell-Seq which combines minimal isothermal amplification with a Phi29 polymerase and a Tn5 transposase that can simultaneously cleave DNA fragments and add adapters for next-generation sequencing, starting with only a single cell. In aim 1 we will develop and optimize the Cell-Seq method. In aim 2 we will validate the method in two clonal cell cultures by comparing the genomes of single cells to million cell samples, to determine error rates and identify potential biases associated with the method. In aim 3 we will apply Cell-Seq to a human breast tumor sample and sequence the genomes of 10 single cells to investigate clonal diversity and genome evolution. We hypothesize that breast tumors are organized into one or more major clonal subpopulations that stably expand to form the tumor mass, not millions of diverse clones, as the prevailing model for tumor progression assumes. We expect that clustering analysis will show evidence for a few major groups, and that within each group genomic mutations will be highly similar. The proposed single-cell sequencing approach is innovative because it can fully resolve heterogeneity in complex populations of cells, whereas standard bulk genomic methods are limited to reporting an average signal. This research is significant because achieving these aims will improve our fundamental understanding of clonal diversity in human breast cancers, and our knowledge of how tumor genomes evolve complex somatic mutations. Our long-term goal is to use single-cell sequencing to study how single cells from human tumors seed metastases and evolve resistance to chemotherapy. Our work is directly aligned with the mission of the NIH to reduce mortality rates in breast cancer through the development of new modalities for early detection and diagnosing heterogeneity in tumors. Our work is also aligned with the goal of the SCAP to cure human diseases through the development of new single-cell genomic technologies. In addition to benefiting the study of cancer, we expect that our tools will have a broad positive impact on many other human diseases, including neurological disorders, immunological diseases, developmental defects and infectious disease.